Research On Building Layout And Target Imaging Method With Multipath Signals

As the good penetration and low loss characteristics of low-frequency electromagnetic(EM)waves,it can be used to reconstruct the building layout and detect concealed targets,which have important applications in the fields of anti-terrorism and stability maintenance,disaster rescue,and urban mapping.However,when the EM waves propagate in the urban environment,due to the occlusion of buildings,they will suffer from reflection,transmission,diffraction,and other complex electromagnetic phenomena,resulting in imaging defocus,target detection and tracking instability.How to make full use of the transmission,reflection and multi-path reflection signals of EM waves to achieve highprecision imaging of building layout and accurate acquisition of target location is a hot spot and difficult research area at home and abroad.This dissertation carries out the research on building layout and target imaging method with multipath signals,and achieves original results in broadband building layout tomographic imaging,joint utilization imaging of transmitted and reflected signals,and Nonline-of-sight(NLOS)layout and target estimation based on multi-path signals,etc.The main work and innovation points are as follows.(1)A two-dimension(2D)tomographic imaging method based on direct wave estimation is proposed.In this method,the linear tomographic model is built by path association and elliptical tomographic projection.Then,a modified total variation method is developed to obtain 2D building layout images with clear edges and complete structure.(2)A tensor constraint-based three-dimensional(3D)building layout enhancement imaging method is proposed under data deficiency conditions.In this method,the prior knowledge of spatial correlation,smoothness and sparsity of height vector group of 3D buildings is found and embedded in the solution framework,which improved 3D layout reconstruction accuracy precision.(3)Considering the limited-angle scanning problem,this dissertation proposes a building layout reconstruction method by jointly utilizing the transmission and reflection signals.The method fuses the reflection projection image and transmission tomographic image by a feature-level detector and shows benefit in solving the problems of wall position offset and serious artifact interference.(4)The dissertation proposes a two-step building layout reconstruction method.In the first step,the position,thickness,and relative permittivity of the external layout are obtained by the wall parameter estimation method.Then,the internal obstacles can be reconstructed by the transmissive data.Finally,the external and internal layouts are fused to generate a panoramic building map.(5)By considering the diffraction and reflections path,a multipath propagation model is build.Then,a NLOS multi-target localization method based on path association is proposed to estimate the non-visual multi-target position under known building layout.Finally,a NLOS parametric multipath propagation model is established,and a shape reshaping sparse reconstruction algorithm is proposed to achieve the joint estimation of NLOS building layout and target position under unknown layout conditions.The above model and method are validated by EM simulation and low-frequency ultra-wideband radar tests,which realize the accurate imaging of urban building layout and targets and lay a theoretical and technical foundation for promoting the detection of complex urban building environments.